Detection of Corn Leaf Diseases Using Convolutional Neural Network With OpenMP Implementation
Dionis A. Padilla, Ramon Alfredo I. Pajes, Jerome T. De Guzman
Abstract
Maize is particularly one of the substantial crop supplies in the Philippines next to rice. The production of the maize crop plays a critical factor in the country's food industry. Disease is one of the major biotic and abiotic constraints to reduce crop yield. Some methods have advantages and disadvantages in contrary to using different accuracies and validity. The main objective of this study is to detect the disease through the leaf in the corn. This paper studied the benefit of both Convolutional Neural Network and OpenMP for disease identification. The study is effective with the objective of identifying and classifying what kind of disease is present in the leaf through the Convolutional Neural Network classifier. Alongside the usage of Convolutional Neural Network and OpenMP implementation, it unites its advantages especially on the sector of execution time rate. The system algorithm was tested using images that were captured for each disease in the corn leaf which was verified by an agriculturist and with the use of Raspberry Pi. As a result, the percent was met with an accuracy of 93%, 89%, and 89% in detecting Leaf Blight, Leaf Rust, and Leaf Spot, respectively. The use of Convolutional Neural Network via OpenMP allowed a high percentage of classification in classifying leaf diseases.